A hybrid approach to implement data driven optimization into production environments

Abstract

The potential of data analytics to improve business processes is commonly recognized. Despite the general enthusiasm, the implementation of data-driven methods in production environments remains low. Although established models, such as CRISP-DM, offer a structured process in order to deploy data analytics in the industry, manufacturing companies still need to choose a starting point, assess the business benefit, and determine a pragmatic course of action. In this paper, we introduce an approach to handle these issues based on a case study from automotive manufacturing. The results are discussed based on a set of requirements derived from the case study.